New CSP research shows how O&M cost savings begin in planning phase
Good decisions in the planning and construction of a CSP plant can bring about lasting operations and maintenance (O&M) cost reductions that boost profitability, several new studies have demonstrated.
If O&M costs decrease by 20%, the internal rate of return of solar tower, parabolic trough, and linear Fresnel plants will increase from 12.33%, 11.72% and 11.43 to 13.41%, 12.79%, and 12.49% respectively, according to one of the studies, a cost-benefit analysis of CSP in China authored by Zhenyu Zhao and Weishang Guo of North China Electric Power University in Beijing.
The O&M costs of CSP plants range from $20/MWh to $40/MWh, the International Renewables Energy Agency (IRENA) has calculated, based on available engineering estimates and recent proposed projects. Because there are no fuel costs, O&M comprises broadly two things: insurance, with an annual cost of around 0.5-1% of the initial capital outlay; and maintenance, which consists largely of cleaning and replacing mirrors.
Therefore, Juan Manuel Medel, Head of O&M at Exera Energia, told New Energy Update, the only way to reduce the cost of O&M is to focus on the “M”, as in maintenance. One way for operators to reduce maintenance costs is by using predictive analytics tools such as the one developed by Exera; another is to design their plant in a way that minimizes cleaning costs.
Size and alignment matter
Future maintenance costs are affected by plant type, and by the size and alignment of heliostats, reflectors or collectors. For example, linear Fresnel reflectors typically use 60-70% of field area compared to about 33% for a parabolic trough system, and the cost of maintaining them is lower due to their easier accessibility, Giovanni Manente of the University of Padova’s Department of Industrial Engineering has noted.
Alignment is important too, as the operators of the 392 MW Ivanpah plant in California found out in 2016 when misaligned heliostats caused a portion of one of the towers to catch on fire, reducing capacity by one-third for more than a month. In renewables, production delays are a non-recoverable loss, and therefore operators must also focus on proactive and predictive maintenance to ensure facilities are in better condition, according to Medel.
Making the right decision on heliostat size can be a major factor in reducing the fixed maintenance costs of a power tower plant, Arvind Sastry Pidaparthi showed in his recent study for Stellenbosch University’s Faculty of Engineering. Based on a cost model developed by the US National Renewable of Energy and using his own model of a 100 MW plant in South Africa with eight hours of energy storage, Pidaparthi concluded that the optimal heliostat is around 115.56m2. He noted in his paper that a 2013 study published by SolarPACES concluded that the optimal heliostat size is 40m2, when taking into account component, installation and checkout, and operations and maintenance.
The NREL cost model found that a plant with 8,709 heliostats of 148m2 area each would have annual maintenance costs of $10.64 million. In Pidaparthi’s model, a field of 8,131 heliostats measured at 115.56m2 each would require the same number of personnel (11) as the larger field, and cost just $8.11 million per year to maintain. Fields of similar overall size but with a higher number of smaller heliostats would require more personnel, and therefore would be more expensive to maintain. For example, a field of 21,670 medium-sized heliostats measured at 43.33m2 each would require 12 personnel and cost $8.27 million annually.
Calculated cleaning routines
Among the largest areas of expenditure in CSP plants are mirror washing, including water costs, and replacement of receivers and mirrors as a result of glass breakage, IRENA noted in its 2017 report on renewable power costs.
The solar specular reflectance in the collector field should naturally be kept at its highest level to ensure high global yield, but keeping the solar collectors clean in an economical manner is the “biggest maintenance challenge” for CSP technology, according to a new report on the maintenance of a pilot plant with parabolic trough system in Louisiana, published last month in the International Journal of Sustainable and Green Energy.
In a previous study of cleaning methods by Spanish researchers, the most effective was found to be deionized water and a brush, which resulted in an average cleaning efficiency of 98.8% in rainy periods and 97.2% in dry periods. The washing procedure employed at the Louisiana plant involved using a pressure washer with deionized water and a microfiber cloth attached to a pole brush designed by 3M. This cleaning method returned the overall reflectivity of the aperture to a value near that of the original performance specification of 95.5%
Cleaning schedule is also crucial, and can be calculated during the planning stage with regular updates during operations, as the authors of the Louisiana study showed. To calculate the most cost-effective schedule, they devised a formula and input the following parameters: the ideal number of days between reflector washing; the cost of cleaning per square meter of surface area; the optical efficiency of the reflectors; the average daily solar energy available per square meter of surface area at the location in question; the soiling rate of the reflector surface as a percentage of the restored reflectivity value; and the energy price in dollars per kilowatt-hour at the specific location.
For the Louisiana site, the optimal interval for washing the collectors was 114 days, or about three times per year. Louisiana, with an electricity price of $0.092/kWh in July 2017, has lower power costs than the Western states where all the commercial-scale CSP plants in the US are located, and therefore, it is cost-effective to wait for longer between washings. Keeping other values constant, California, with a price of $0.177/kWh, would have a washing interval of 82 days to minimize the cost.
Although the results are very site-specific, the same method could be used to calculate the optimal cleaning schedule for other sites, the authors said.